In [1]:
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
In [2]:
df = pd.read_csv('E:/Data Science/Projects/Python_projects/Loan_data_analysis_project/Loan_dataset.csv')
In [3]:
df
Out[3]:
Loan ID Customer ID Loan Status Current Loan Amount Term Credit Score Annual Income Years in current job Home Ownership Purpose Monthly Debt Years of Credit History Months since last delinquent Number of Open Accounts Number of Credit Problems Current Credit Balance Maximum Open Credit Bankruptcies Tax Liens
0 14dd8831-6af5-400b-83ec-68e61888a048 981165ec-3274-42f5-a3b4-d104041a9ca9 Fully Paid 445412.0 Short Term 709.0 1167493.0 8 years Home Mortgage Home Improvements 5214.74 17.2 NaN 6.0 1.0 228190.0 416746.0 1.0 0.0
1 4771cc26-131a-45db-b5aa-537ea4ba5342 2de017a3-2e01-49cb-a581-08169e83be29 Fully Paid 262328.0 Short Term NaN NaN 10+ years Home Mortgage Debt Consolidation 33295.98 21.1 8.0 35.0 0.0 229976.0 850784.0 0.0 0.0
2 4eed4e6a-aa2f-4c91-8651-ce984ee8fb26 5efb2b2b-bf11-4dfd-a572-3761a2694725 Fully Paid 99999999.0 Short Term 741.0 2231892.0 8 years Own Home Debt Consolidation 29200.53 14.9 29.0 18.0 1.0 297996.0 750090.0 0.0 0.0
3 77598f7b-32e7-4e3b-a6e5-06ba0d98fe8a e777faab-98ae-45af-9a86-7ce5b33b1011 Fully Paid 347666.0 Long Term 721.0 806949.0 3 years Own Home Debt Consolidation 8741.90 12.0 NaN 9.0 0.0 256329.0 386958.0 0.0 0.0
4 d4062e70-befa-4995-8643-a0de73938182 81536ad9-5ccf-4eb8-befb-47a4d608658e Fully Paid 176220.0 Short Term NaN NaN 5 years Rent Debt Consolidation 20639.70 6.1 NaN 15.0 0.0 253460.0 427174.0 0.0 0.0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
100509 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
100510 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
100511 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
100512 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
100513 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN

100514 rows × 19 columns

In [4]:
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 100514 entries, 0 to 100513
Data columns (total 19 columns):
 #   Column                        Non-Null Count   Dtype  
---  ------                        --------------   -----  
 0   Loan ID                       100000 non-null  object 
 1   Customer ID                   100000 non-null  object 
 2   Loan Status                   100000 non-null  object 
 3   Current Loan Amount           100000 non-null  float64
 4   Term                          100000 non-null  object 
 5   Credit Score                  80846 non-null   float64
 6   Annual Income                 80846 non-null   float64
 7   Years in current job          95778 non-null   object 
 8   Home Ownership                100000 non-null  object 
 9   Purpose                       100000 non-null  object 
 10  Monthly Debt                  100000 non-null  float64
 11  Years of Credit History       100000 non-null  float64
 12  Months since last delinquent  46859 non-null   float64
 13  Number of Open Accounts       100000 non-null  float64
 14  Number of Credit Problems     100000 non-null  float64
 15  Current Credit Balance        100000 non-null  float64
 16  Maximum Open Credit           99998 non-null   float64
 17  Bankruptcies                  99796 non-null   float64
 18  Tax Liens                     99990 non-null   float64
dtypes: float64(12), object(7)
memory usage: 14.6+ MB
In [5]:
df.columns
Out[5]:
Index(['Loan ID', 'Customer ID', 'Loan Status', 'Current Loan Amount', 'Term',
       'Credit Score', 'Annual Income', 'Years in current job',
       'Home Ownership', 'Purpose', 'Monthly Debt', 'Years of Credit History',
       'Months since last delinquent', 'Number of Open Accounts',
       'Number of Credit Problems', 'Current Credit Balance',
       'Maximum Open Credit', 'Bankruptcies', 'Tax Liens'],
      dtype='object')

1) What is the average current loan amount for each loan status?¶

In [7]:
df['Current Loan Amount'].unique()
Out[7]:
array([  445412.,   262328., 99999999., ...,   100254.,   274076.,
             nan])
In [8]:
value_counts = df['Current Loan Amount'].value_counts()
In [9]:
sorted_value_counts = value_counts.sort_index()     

# Sorting Values
In [10]:
print(pd.DataFrame(sorted_value_counts))

                     count
Current Loan Amount       
10802.0                  1
11242.0                  1
15422.0                  2
21098.0                  1
21450.0                  3
...                    ...
789030.0                 9
789096.0                16
789184.0                 6
789250.0                 3
99999999.0           11484

[22004 rows x 1 columns]

The value "99999999" appears a lot of times in the data.¶

It is a huge number, plus lot of people having the same number would mean either they have a huge loan pending or they paid it off¶

First, we check whether this amount an outstanding amount or a paid amount¶

In [12]:
condition = df['Current Loan Amount'] == 99999999
In [13]:
filtered_loans = df[condition][['Current Loan Amount','Loan Status']]
In [14]:
filtered_loans['Current Loan Amount'].unique()
Out[14]:
array([99999999.])
In [15]:
filtered_loans['Loan Status'].unique()
Out[15]:
array(['Fully Paid'], dtype=object)

As we deduced, all the "99999999" amount is Fully Paid¶

Hence, we conclude this as an amount that is already paid¶

and this amount has to be '0' rather than '99999999'¶

In [17]:
# Replacing '99999999' with 0

df = df.replace(99999999,0)
In [18]:
df['Current Loan Amount'].unique()
Out[18]:
array([445412., 262328.,      0., ..., 100254., 274076.,     nan])
In [19]:
df['Current Loan Amount'].isna().unique()     

# To Check null values
Out[19]:
array([False,  True])
In [20]:
df1= df.dropna(subset='Current Loan Amount')  

# Creating a New DataFrame and removing all the null values
In [21]:
df1['Current Loan Amount'].isna().unique()     

# To Recheck null values
Out[21]:
array([False])
In [22]:
df1 = df1[df1['Current Loan Amount'] != 0]     

# Removing all the values that are "Fully Paid"

Numeric Representation¶

In [24]:
pd.DataFrame(df1['Current Loan Amount'].groupby(by=df1['Loan Status']).mean().round(2))
Out[24]:
Current Loan Amount
Loan Status
Charged Off 324975.23
Fully Paid 307962.36

Graphical Representation¶

In [26]:
Average_Loan_Amounts = df1.groupby('Loan Status')['Current Loan Amount'].mean().round(2).reset_index()

# To Create a DataFrame
In [27]:
Average_Loan_Amounts
Out[27]:
Loan Status Current Loan Amount
0 Charged Off 324975.23
1 Fully Paid 307962.36
In [28]:
plt.figure(figsize=(10,6))
plt.bar(data=Average_Loan_Amounts, x='Loan Status', height='Current Loan Amount')
plt.xlabel('Loan Status')
plt.ylabel('Avg Current Loan Amount')
plt.grid(axis='y')  # Grid lines for better readability
plt.title('Average Current Loan Amount for each Loan Status')
plt.show()
No description has been provided for this image

2) How does the credit score vary with the annual income?¶

In [30]:
df['Credit Score'].unique() 

# Checking values
Out[30]:
array([ 709.,   nan,  741.,  721., 7290.,  730.,  678.,  739.,  728.,
        740.,  743.,  727.,  723.,  747.,  687.,  750.,  714.,  724.,
        704.,  688.,  749.,  746.,  737.,  729.,  733.,  725.,  745.,
        720.,  718.,  682., 7120.,  680.,  710.,  598.,  719., 6610.,
        652.,  736., 7380.,  644.,  672., 7370.,  699.,  751.,  694.,
        675.,  657.,  748.,  666.,  734.,  742.,  705.,  731., 6240.,
        712.,  685.,  717.,  722.,  618.,  676.,  692., 7210.,  732.,
        649.,  695.,  744.,  686.,  637.,  697.,  706.,  715.,  707.,
        726.,  738., 7500.,  716., 7020.,  651.,  708.,  698.,  689.,
        735.,  703.,  693., 7080.,  645.,  691.,  673.,  700., 7140.,
        658.,  674.,  654., 7490.,  681.,  696.,  713.,  668., 7360.,
        659., 7160.,  647.,  683.,  670.,  623.,  711.,  639.,  671.,
       6990.,  614.,  667.,  701., 7410., 7310., 7040., 7060.,  615.,
       7200.,  656.,  653.,  664.,  613., 7450., 7300.,  636.,  702.,
       7430.,  594., 7170.,  646., 7000.,  669., 7280., 7090., 6970.,
       7460., 6920.,  612., 7350., 7190.,  690., 6690., 6750.,  641.,
        665.,  677.,  684.,  655., 7010., 6530.,  630., 7270.,  663.,
        638.,  596., 7420.,  660., 6810.,  679., 7070., 7390., 7470.,
        609.,  603., 6670., 6710., 6580., 6940., 6800., 7110.,  661.,
       7130.,  643.,  633., 5860., 6960.,  629.,  648., 7150., 6400.,
       6830., 6360., 6650.,  625., 6740.,  635., 7260., 7480., 6980.,
        662.,  640., 7250., 7320.,  627., 7180., 6640., 6780.,  611.,
       6090.,  604.,  642., 7240.,  621., 6550., 6900., 6590., 6840.,
       6660., 7400.,  632., 6760., 7340., 6860., 6870., 6680.,  602.,
       7220.,  616.,  600., 6460., 6500.,  597., 6880., 6540.,  622.,
       6100., 7330.,  619.,  624., 7510., 6930., 6430., 6160.,  628.,
        599., 7030., 6630., 7230.,  606., 6890., 6820., 6850., 7100.,
       6210., 6150., 7440., 6080.,  650., 6200., 6570.,  626., 6110.,
       6560., 6730., 6410.,  605., 6310., 5960.,  589., 6350.,  607.,
       6470., 6520., 6700.,  608.,  631., 6950., 6720., 6290., 6770.,
        620., 6480., 6380., 7050., 6390.,  590.,  634., 6910., 6620.,
       6370., 6600., 6250.,  587.,  588.,  591.,  617.,  610., 6450.,
       6010.,  595.,  601.,  586., 6490., 6510., 5850.,  585., 6300.,
       6790., 6330., 6130., 6220., 6440., 6340., 6270., 6260., 6170.,
       5940., 5950., 6420., 6180.,  592., 5900., 6320., 6230., 6120.,
       6280.,  593., 5920., 6140., 6070., 6190., 5930., 6060., 5890.,
       6050.])

The highest credit score for the US is 850.¶

Some numbers above 1000. But all of them have one thing in common.¶

In [32]:
filtered_values = df['Credit Score'][df['Credit Score']>1000]
In [33]:
filtered_values
Out[33]:
5        7290.0
43       7120.0
51       6610.0
56       7380.0
65       7370.0
          ...  
99851    7420.0
99872    7070.0
99906    6880.0
99944    6790.0
99947    6940.0
Name: Credit Score, Length: 4551, dtype: float64
In [34]:
sorted_values = filtered_values.sort_values()
In [35]:
sorted_values.astype('int').unique()
Out[35]:
array([5850, 5860, 5890, 5900, 5920, 5930, 5940, 5950, 5960, 6010, 6050,
       6060, 6070, 6080, 6090, 6100, 6110, 6120, 6130, 6140, 6150, 6160,
       6170, 6180, 6190, 6200, 6210, 6220, 6230, 6240, 6250, 6260, 6270,
       6280, 6290, 6300, 6310, 6320, 6330, 6340, 6350, 6360, 6370, 6380,
       6390, 6400, 6410, 6420, 6430, 6440, 6450, 6460, 6470, 6480, 6490,
       6500, 6510, 6520, 6530, 6540, 6550, 6560, 6570, 6580, 6590, 6600,
       6610, 6620, 6630, 6640, 6650, 6660, 6670, 6680, 6690, 6700, 6710,
       6720, 6730, 6740, 6750, 6760, 6770, 6780, 6790, 6800, 6810, 6820,
       6830, 6840, 6850, 6860, 6870, 6880, 6890, 6900, 6910, 6920, 6930,
       6940, 6950, 6960, 6970, 6980, 6990, 7000, 7010, 7020, 7030, 7040,
       7050, 7060, 7070, 7080, 7090, 7100, 7110, 7120, 7130, 7140, 7150,
       7160, 7170, 7180, 7190, 7200, 7210, 7220, 7230, 7240, 7250, 7260,
       7270, 7280, 7290, 7300, 7310, 7320, 7330, 7340, 7350, 7360, 7370,
       7380, 7390, 7400, 7410, 7420, 7430, 7440, 7450, 7460, 7470, 7480,
       7490, 7500, 7510])

They are divisible by 10 and less than 850.¶

So there most likely is a data error.¶

Hence assuming this, we divide those numbers by 10.¶

In [37]:
Col_to_mod = 'Credit Score'
In [38]:
df.loc[df[Col_to_mod]>1000,Col_to_mod]/=10  

# Divide only the numbers > 1000 in the specified column by 10
In [39]:
df['Credit Score'].unique()  # To check data and null values
Out[39]:
array([709.,  nan, 741., 721., 729., 730., 678., 739., 728., 740., 743.,
       727., 723., 747., 687., 750., 714., 724., 704., 688., 749., 746.,
       737., 733., 725., 745., 720., 718., 682., 712., 680., 710., 598.,
       719., 661., 652., 736., 738., 644., 672., 699., 751., 694., 675.,
       657., 748., 666., 734., 742., 705., 731., 624., 685., 717., 722.,
       618., 676., 692., 732., 649., 695., 744., 686., 637., 697., 706.,
       715., 707., 726., 716., 702., 651., 708., 698., 689., 735., 703.,
       693., 645., 691., 673., 700., 658., 674., 654., 681., 696., 713.,
       668., 659., 647., 683., 670., 623., 711., 639., 671., 614., 667.,
       701., 615., 656., 653., 664., 613., 636., 594., 646., 669., 612.,
       690., 641., 665., 677., 684., 655., 630., 663., 638., 596., 660.,
       679., 609., 603., 643., 633., 586., 629., 648., 640., 625., 635.,
       662., 627., 611., 604., 642., 621., 632., 602., 616., 600., 650.,
       597., 622., 610., 619., 628., 599., 606., 608., 620., 626., 605.,
       631., 589., 607., 590., 634., 587., 588., 591., 617., 601., 595.,
       585., 592., 593.])
In [40]:
df2= df.dropna(subset='Credit Score')  

# Creating a New DataFrame and removing all the null values
In [41]:
df2['Credit Score'].unique()

# To check null values as well as updated values
Out[41]:
array([709., 741., 721., 729., 730., 678., 739., 728., 740., 743., 727.,
       723., 747., 687., 750., 714., 724., 704., 688., 749., 746., 737.,
       733., 725., 745., 720., 718., 682., 712., 680., 710., 598., 719.,
       661., 652., 736., 738., 644., 672., 699., 751., 694., 675., 657.,
       748., 666., 734., 742., 705., 731., 624., 685., 717., 722., 618.,
       676., 692., 732., 649., 695., 744., 686., 637., 697., 706., 715.,
       707., 726., 716., 702., 651., 708., 698., 689., 735., 703., 693.,
       645., 691., 673., 700., 658., 674., 654., 681., 696., 713., 668.,
       659., 647., 683., 670., 623., 711., 639., 671., 614., 667., 701.,
       615., 656., 653., 664., 613., 636., 594., 646., 669., 612., 690.,
       641., 665., 677., 684., 655., 630., 663., 638., 596., 660., 679.,
       609., 603., 643., 633., 586., 629., 648., 640., 625., 635., 662.,
       627., 611., 604., 642., 621., 632., 602., 616., 600., 650., 597.,
       622., 610., 619., 628., 599., 606., 608., 620., 626., 605., 631.,
       589., 607., 590., 634., 587., 588., 591., 617., 601., 595., 585.,
       592., 593.])
In [42]:
df2['Annual Income'].isna().unique()     

# To Check null values
Out[42]:
array([False])
In [43]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Annual Income',y='Credit Score')
plt.grid(axis='x')  # Grid Lines for better readability
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Credit Score')
plt.title('Annual Income vs Credit Score')
plt.show()
No description has been provided for this image
In [44]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Annual Income',y='Credit Score')
plt.xlim(-200000,30000000)     # To not include outliers
plt.xlabel('Annual Income (per 10 Million USD)') 
plt.ylabel('Credit Score')
plt.grid(axis='x')  # Grid Lines for better readability
plt.title('Annual Income vs Credit Score')
plt.show()
No description has been provided for this image
In [45]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Annual Income',y='Credit Score')
plt.xlim(-200000,15000000)     # To not include outliers
plt.xlabel('Annual Income (per 10 Million USD)') 
plt.ylabel('Credit Score')
plt.grid(axis='x')  # Grid Lines for better readability
plt.title('Annual Income vs Credit Score')
plt.show()
No description has been provided for this image
In [46]:
df2[['Credit Score','Annual Income']].corr()
Out[46]:
Credit Score Annual Income
Credit Score 1.000000 -0.005092
Annual Income -0.005092 1.000000
In [47]:
pd.DataFrame(df2[['Credit Score','Annual Income']].var().round(2))
Out[47]:
0
Credit Score 8.007300e+02
Annual Income 1.169340e+12
In [48]:
pd.DataFrame(df2[['Credit Score','Annual Income']].std().round(2))
Out[48]:
0
Credit Score 28.3
Annual Income 1081360.2
In [49]:
df2 = df2[df2['Annual Income'] < 10000000]   

# To remove Outliers
In [50]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Annual Income',y='Credit Score')
plt.grid(axis='x')  # Grid Lines for better readability
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Credit Score')
plt.title('Annual Income vs Credit Score')
plt.show()
No description has been provided for this image
In [51]:
df2[['Credit Score','Annual Income']].corr()
Out[51]:
Credit Score Annual Income
Credit Score 1.000000 -0.008833
Annual Income -0.008833 1.000000
In [52]:
pd.DataFrame(df2[['Credit Score','Annual Income']].var().round(2))
Out[52]:
0
Credit Score 8.005300e+02
Annual Income 6.649314e+11
In [53]:
pd.DataFrame(df2[['Credit Score','Annual Income']].std().round(2))
Out[53]:
0
Credit Score 28.29
Annual Income 815433.24
In [54]:
df2= df.dropna(subset='Credit Score')  

# Reverting this Dataframe for further use

3) Is there a correlation between the number of open accounts and the current credit balance?¶

In [56]:
df['Number of Open Accounts'].isna().unique()    

# To check null values
Out[56]:
array([False,  True])
In [57]:
df3 = df.dropna(subset='Number of Open Accounts')

# Creating a New DataFrame and removing all the null values
In [58]:
df3['Number of Open Accounts'].isna().unique()   

# To recheck null values
Out[58]:
array([False])
In [59]:
df3['Current Credit Balance'].isna().unique()  

# To check null values
Out[59]:
array([False])
In [60]:
df3['Current Credit Balance'].max()    

# To determine the range
Out[60]:
32878968.0
In [61]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df3,x='Current Credit Balance',y='Number of Open Accounts')
plt.xlabel('Current Credit Balance (Per 10 Million USD)')
plt.ylabel('Number of Open Accounts')
plt.grid(axis='x')  # Grid Lines for better readability
plt.title('Relationship between Current Credit Balance & No. of Open Accounts')
plt.show()
No description has been provided for this image
In [62]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df3,x='Current Credit Balance',y='Number of Open Accounts')
plt.xlim(-100000,10000000)  # To not include outliers
plt.xlabel('Current Credit Balance (Per 10 Million USD)')
plt.ylabel('Number of Open Accounts')
plt.grid(axis='x')  # Grid Lines for better readability
plt.title('Relationship between Current Credit Balance & No. of Open Accounts')
plt.show()
No description has been provided for this image
In [63]:
df3[['Number of Open Accounts','Current Credit Balance']].corr()
Out[63]:
Number of Open Accounts Current Credit Balance
Number of Open Accounts 1.000000 0.228136
Current Credit Balance 0.228136 1.000000
In [64]:
df3 = df3[df3['Current Credit Balance'] < 10000000]   

# To remove Outliers
In [65]:
plt.figure(figsize=(10,6))
sns.scatterplot(x='Current Credit Balance',y='Number of Open Accounts',data=df3)
plt.xlabel('Current Credit Balance (per Million USD)')
plt.ylabel('Number of Open Accounts')
plt.grid(axis='x')  # Grid Lines for better readability
plt.title('Relationship between Current Credit Balance & No. of Open Accounts')
plt.show()
No description has been provided for this image
In [66]:
df3[['Number of Open Accounts','Current Credit Balance']].corr()
Out[66]:
Number of Open Accounts Current Credit Balance
Number of Open Accounts 1.000000 0.239696
Current Credit Balance 0.239696 1.000000

As the Correlation between Number of Open Accounts and Current Credit Balance is 0.239696, the correlation between them is on a weaker side¶

4) What is the distribution of credit scores across different home ownership types?¶

In [69]:
df2['Credit Score'].isna().unique()

# Using already made up DataFrame and checking null values
Out[69]:
array([False])
In [70]:
df2['Home Ownership'].isna().unique()     

# Using already made up DataFrame and checking null values
Out[70]:
array([False])

Graphical Representation¶

In [72]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df2,x='Credit Score',y='Home Ownership')
plt.xlabel('Credit Score')
plt.ylabel('Home Ownership')
plt.grid(axis='x')  # Grid lines for better readability
plt.title('Distribution of Credit Score across Home Ownership Types')
plt.show()
No description has been provided for this image

Numeric Representation¶

In [74]:
df2.groupby('Home Ownership')['Credit Score'].describe()
Out[74]:
count mean std min 25% 50% 75% max
Home Ownership
HaveMortgage 185.0 716.383784 22.028854 651.0 702.0 719.0 733.0 751.0
Home Mortgage 39089.0 717.526568 28.540485 585.0 704.0 724.0 740.0 751.0
Own Home 7480.0 715.346257 28.753753 585.0 702.0 721.0 738.0 751.0
Rent 34092.0 715.086912 27.884310 585.0 702.0 721.0 737.0 751.0

5) How does the annual income differ for different purposes of loans?¶

In [76]:
df4 = pd.DataFrame(df.groupby(by='Purpose')['Annual Income'].mean().round(2))

Numeric Representation¶

In [78]:
df4
Out[78]:
Annual Income
Purpose
Business Loan 1591962.79
Buy House 1495251.47
Buy a Car 1220352.74
Debt Consolidation 1362463.57
Educational Expenses 1071051.32
Home Improvements 1667437.36
Medical Bills 1409286.17
Other 1348516.49
Take a Trip 1219898.67
major_purchase 1430753.50
moving 1591375.25
other 1286008.49
renewable_energy 1473739.75
small_business 1547456.22
vacation 1155289.53
wedding 1464232.62

Graphical Representation¶

In [80]:
plt.figure(figsize=(10,6))
sns.barplot(data=df4,x='Annual Income',y='Purpose')
plt.xlabel('Annual Income (Per 10 Million Dollars)')
plt.ylabel('Purpose')
plt.grid(axis='x')     # Grid Lines for better readability
plt.title('Purpose of Loans compared to Annual Income')
plt.show()
No description has been provided for this image

6) What is the average monthly debt for each term (short-term vs. long-term)?¶

In [82]:
df5 = df.dropna(subset='Term')

# Creating a New DataFrame and removing all the null values
In [83]:
df5['Monthly Debt'].isna().unique()

# To check null values
Out[83]:
array([False])
In [84]:
df5['Term'].isna().unique()

# To check null values
Out[84]:
array([False])
In [85]:
df5 = pd.DataFrame(df5.groupby(by='Term')['Monthly Debt'].mean().round(2))

Numeric Representation¶

In [87]:
df5
Out[87]:
Monthly Debt
Term
Long Term 21585.66
Short Term 17274.16

Graphical Representation¶

In [89]:
plt.figure(figsize=(10,6))
sns.barplot(data=df5,x='Term',y='Monthly Debt')
plt.xlabel('Term')
plt.ylabel('Average Monthly Debt')
plt.grid(axis='y')    # Grid Lines for better readability  
plt.title('Average Monthly Debt for Each Term')
plt.show()
No description has been provided for this image

7) Is there a correlation between years of credit history and the current credit balance?¶

In [91]:
df6 = df.dropna(subset='Current Credit Balance')

# Creating a New DataFrame and removing all the null values
In [92]:
df6['Current Credit Balance'].isna().unique()     

# To check null values
Out[92]:
array([False])
In [93]:
df6['Years of Credit History'].unique()

# To check null values
Out[93]:
array([17.2, 21.1, 14.9, 12. ,  6.1, 17.3, 19.6,  8.2, 22.6, 13.9, 16. ,
       23.2,  8.5, 13.3, 24.7, 19.4, 20.5, 22. , 27.1, 22.7, 17.4, 11.9,
       15.7, 17.5, 14.4, 22.3, 14. , 20. , 43.7, 18. , 15.4, 14.6, 24.5,
       14.1, 20.6, 28.8, 16.6, 16.7, 14.8, 14.5, 21.6, 18.6, 17.8, 20.9,
       17. , 15.6, 36.6, 40.8, 11.4,  9.2, 16.1, 15. , 18.3, 12.1, 31.5,
       19.3, 21.8, 15.1, 25.6, 31.4, 29.5, 13.8, 21.5,  6.4, 10.5, 13. ,
       19. , 11.8, 28.2, 12.2, 44.5, 27.4, 20.4, 22.4, 29.1, 31.3, 17.9,
       14.2, 25.9, 14.7, 23.3, 19.7, 10. , 19.5, 11. , 12.3, 13.6, 26.5,
       10.2, 16.9,  7. , 12.5, 15.9, 10.7, 13.4, 20.7,  9. , 23.4,  9.5,
       34.4, 18.4, 23.5, 14.3, 50.1, 13.2, 33.1, 23.9, 13.5, 43.3, 24. ,
       23. , 16.2, 29.2, 18.2, 44. , 21. , 30.5,  7.1,  6.6, 15.8, 16.5,
       17.6, 13.7, 36.3, 22.5,  8.6, 15.5, 10.8, 18.7, 30.9, 17.1, 16.3,
        4.9, 21.4, 11.1, 11.7, 10.3, 18.5,  8.7,  9.9, 19.2, 21.3, 22.2,
       33.7, 27. , 18.8, 13.1, 15.2, 15.3, 21.2, 20.8,  9.7, 11.5,  9.3,
       19.9, 42.4, 22.8, 35. , 24.1, 36.4, 30.6, 12.8, 23.8, 25.5, 38.5,
       24.4, 12.4, 10.1, 12.9, 18.1, 11.3,  7.5, 16.4, 20.2,  9.8, 23.6,
       27.2, 32.9,  8.9, 16.8, 10.6, 26.9, 11.2, 40.3, 20.1,  8. , 28.4,
       21.9, 21.7, 48.7, 32.5,  8.8, 30. , 22.9, 26. , 35.6, 28.9, 12.6,
       38. , 26.3, 28.5, 31.2, 26.4,  8.4, 35.9, 29. ,  8.3, 24.3, 25.7,
       33.5, 33.4, 19.8, 31. , 19.1, 30.8, 27.7, 32.2, 38.3, 32.4, 29.9,
       22.1, 26.7, 29.8,  9.4, 27.5, 27.8,  7.7, 25. ,  9.6, 10.9, 17.7,
       31.8, 12.7, 27.6, 37.1, 37.4, 24.2, 31.7, 25.2, 34.2, 18.9,  6.8,
       32.6, 27.9, 26.1, 32.3,  7.6, 32.1,  5. , 23.1,  5.8,  5.5, 34.1,
       24.6, 34. ,  7.9, 10.4, 28.1, 20.3, 24.9, 39. , 25.8, 30.4, 35.4,
        7.8, 28.3, 34.3, 28.6, 38.9, 39.4, 11.6, 32.7,  9.1, 26.2,  6.2,
        5.4, 40.1, 28.7, 45.3, 36. , 29.3, 34.5,  7.4, 33. , 37.8,  8.1,
       42.3, 25.4, 26.8,  5.7,  6.5,  6. , 39.6, 38.8, 29.7,  6.9, 30.3,
        6.7, 30.2, 28. , 26.6, 31.9, 39.9, 24.8, 37. , 23.7,  4.5, 35.5,
       46.2, 42.7, 32.8, 46.9,  4.2, 41.5, 38.7, 29.4, 30.7, 33.9,  7.3,
        5.3, 43.5,  5.2, 35.2, 38.6,  3.7, 37.5, 31.1, 35.8, 33.3,  6.3,
       31.6, 39.1, 30.1, 43.2, 40.2, 40.7, 32. , 36.2, 47. , 25.3, 34.9,
        5.9, 34.7, 48.9, 33.6, 39.5, 29.6,  4.4,  7.2, 46.3, 42.8, 25.1,
       40.4, 38.1, 42. , 46. , 41. , 41.1, 44.6, 42.5, 33.2, 50.9, 41.2,
       34.8, 48.8, 43.8, 36.8, 33.8, 41.9, 27.3, 43.9, 46.8, 41.3, 41.6,
       37.2, 48.2, 37.6, 42.6, 41.7, 39.3, 46.1, 46.7, 41.4,  3.6, 38.2,
       36.5, 40. , 40.5, 36.7,  4.8, 52.8, 42.1, 40.6, 35.3,  5.6, 46.5,
       37.3, 53.2, 45.5, 42.9, 38.4, 37.7, 44.3,  5.1, 35.7, 34.6, 39.7,
       51.4,  4.3, 36.9, 35.1, 47.5, 41.8, 39.2, 47.4, 43.4,  4. , 39.8,
       44.1, 43. , 37.9, 44.4, 36.1, 48.3, 51. , 47.3,  4.1, 43.6, 45. ,
       49.9, 49.6, 49. ,  4.6, 56.5, 43.1, 45.1, 45.6, 65. , 47.1, 49.4,
       51.5, 54. , 46.4,  3.9,  4.7, 47.9, 45.8, 52.9, 55.8, 52.5, 49.5,
       44.7, 51.9, 40.9,  3.8, 50. , 44.8, 48.5, 45.4, 70.5, 42.2, 59.5,
       56.8, 49.3, 52.6, 48.6, 51.6, 52.3, 59.7, 44.2, 51.8, 50.3, 54.5,
       48. , 45.2, 47.8, 49.1, 60.5, 50.6, 47.7, 45.9, 58. , 46.6, 50.4,
       50.5, 50.8, 53.4, 57.7, 45.7, 53.8, 51.7, 47.6, 59.9, 53.1, 54.2,
       52. , 57. , 48.4, 57.5, 49.2, 56. , 50.2, 49.8, 49.7, 52.2, 51.3])
In [94]:
df6['Years of Credit History'].isna().unique()     

# To check null values
Out[94]:
array([False])
In [95]:
df6['Current Credit Balance'].max()

# To check the range
Out[95]:
32878968.0
In [96]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df6,x='Current Credit Balance',y='Years of Credit History')
plt.grid(axis='x')   # Grid Lines for better readability
plt.xlabel('Current Credit Balance (per Million USD)')
plt.ylabel('Years of Credit History')
plt.title('Correlation between Current Credit Balance & Years of Credit History')
plt.show()
No description has been provided for this image
In [97]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df6,x='Current Credit Balance',y='Years of Credit History')
plt.grid(axis='x')          # Grid Lines for better readability
plt.xlim(-200000,10000000)     # To not include outliers
plt.xlabel('Current Credit Balance (per Million USD)')
plt.ylabel('Years of Credit History')
plt.title('Correlation between Current Credit Balance & Years of Credit History')
plt.show()
No description has been provided for this image
In [98]:
df6[['Years of Credit History','Current Credit Balance']].corr()
Out[98]:
Years of Credit History Current Credit Balance
Years of Credit History 1.00000 0.20847
Current Credit Balance 0.20847 1.00000
In [99]:
df6['Current Credit Balance'].max()
Out[99]:
32878968.0
In [100]:
df6 = df6[df6['Current Credit Balance'] < 1000000]   

# To remove Outliers
In [101]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df6,x='Current Credit Balance',y='Years of Credit History')
plt.grid(axis='x')    # Grid Lines for better readability
plt.xlabel('Current Credit Balance (per Million USD)')
plt.ylabel('Years of Credit History')
plt.title('Correlation between Current Credit Balance & Years of Credit History')
plt.show()
No description has been provided for this image
In [102]:
df6[['Years of Credit History','Current Credit Balance']].corr()
Out[102]:
Years of Credit History Current Credit Balance
Years of Credit History 1.000000 0.237888
Current Credit Balance 0.237888 1.000000

As the Correlation between Number of Open Accounts and Current Credit Balance is 0.237888, the correlation between them is on a weaker side.¶

Hence, there is no correlation between them.¶

8) How does the credit score vary with the years in the current job?¶

In [106]:
plt.figure(figsize=(10, 10))
sns.scatterplot(data=df,x='Credit Score',y='Years in current job')
plt.title('Credit Score vs Years in Current Job')
plt.show()
No description has been provided for this image
In [107]:
df['Years in current job'].unique()          

# To check all Unique Values
Out[107]:
array(['8 years', '10+ years', '3 years', '5 years', '< 1 year',
       '2 years', '4 years', '9 years', '7 years', '1 year', nan,
       '6 years'], dtype=object)
In [108]:
df7 = df.dropna(subset='Years in current job')   

# Removing null Values
In [109]:
df7['Years in current job'].unique()            

# To recheck null Values
Out[109]:
array(['8 years', '10+ years', '3 years', '5 years', '< 1 year',
       '2 years', '4 years', '9 years', '7 years', '1 year', '6 years'],
      dtype=object)
In [110]:
# Function to convert the strings to integers

def convert_to_integers(year_str):
    if '<' in year_str:
        return 0              # Assuming "<1 year" means 0 years
    elif '+' in year_str:
        return int(year_str.split('+')[0])    # Take the numeric part before "+"
    else: 
        return int(year_str.split()[0])       # Split by space and take the first part
In [111]:
# Apply the conversion function to the "Years in current job" column

df7['Years in current job'] = df7['Years in current job'].apply(convert_to_integers)

C:\Users\Admin\AppData\Local\Temp\ipykernel_24012\2970841671.py:3: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df7['Years in current job'] = df7['Years in current job'].apply(convert_to_integers)
In [112]:
# To Check whether the conversion was successful

df7['Years in current job'].unique()
Out[112]:
array([ 8, 10,  3,  5,  0,  2,  4,  9,  7,  1,  6], dtype=int64)
In [113]:
# To check Null Values

df7['Credit Score'].unique()
Out[113]:
array([709.,  nan, 741., 721., 729., 730., 678., 739., 728., 740., 743.,
       727., 723., 747., 687., 750., 714., 724., 704., 688., 749., 746.,
       737., 733., 725., 745., 720., 718., 682., 712., 680., 710., 598.,
       719., 661., 652., 736., 738., 644., 672., 699., 751., 694., 657.,
       748., 666., 734., 742., 705., 624., 685., 717., 722., 618., 676.,
       692., 732., 649., 695., 744., 686., 637., 697., 706., 715., 707.,
       726., 716., 702., 651., 708., 698., 689., 703., 693., 645., 691.,
       700., 731., 658., 735., 674., 654., 681., 696., 713., 668., 659.,
       647., 683., 675., 670., 623., 711., 639., 671., 614., 667., 615.,
       656., 653., 664., 613., 636., 594., 646., 669., 701., 612., 673.,
       690., 641., 665., 677., 684., 655., 630., 663., 638., 596., 660.,
       679., 609., 603., 643., 633., 586., 629., 648., 640., 625., 635.,
       662., 627., 611., 604., 642., 621., 632., 602., 616., 600., 650.,
       597., 622., 610., 619., 628., 599., 606., 608., 620., 626., 605.,
       631., 589., 607., 590., 634., 587., 588., 591., 617., 601., 595.,
       585., 592., 593.])
In [114]:
# Removing Null Values

df7 = df7.dropna(subset='Credit Score')
In [115]:
plt.figure(figsize=(10, 10))
sns.scatterplot(data=df7,x='Credit Score',y='Years in current job')
plt.xlabel('Years in Current Job')
plt.title('Credit Score vs Years in Current Job')
plt.show()
print('0 in Y-axis denotes less than 1 year and 10 denotes more than 10 years ')
No description has been provided for this image 
0 in Y-axis denotes less than 1 year and 10 denotes more than 10 years
In [116]:
df7[['Credit Score','Years in current job']].corr()
Out[116]:
Credit Score Years in current job
Credit Score 1.000000 -0.017076
Years in current job -0.017076 1.000000
In [117]:
pd.DataFrame(df7.groupby('Years in current job')['Credit Score'].var().round(2))
Out[117]:
Credit Score
Years in current job
0 821.98
1 763.28
2 782.46
3 767.91
4 761.60
5 803.63
6 812.67
7 837.17
8 839.64
9 791.95
10 817.69
In [118]:
pd.DataFrame(df7.groupby('Years in current job')['Credit Score'].std().round(2))
Out[118]:
Credit Score
Years in current job
0 28.67
1 27.63
2 27.97
3 27.71
4 27.60
5 28.35
6 28.51
7 28.93
8 28.98
9 28.14
10 28.60

9) What is the relationship between the number of credit problems and the number of open accounts?¶

In [120]:
df8 = df.dropna(subset='Number of Open Accounts')

# Creating a New DataFrame and removing all the null values
In [121]:
df8['Number of Open Accounts'].isna().unique()

# To check null values
Out[121]:
array([False])
In [122]:
df8['Number of Credit Problems'].isna().unique()

# To check null values
Out[122]:
array([False])
In [123]:
plt.figure(figsize=(10, 6))
sns.scatterplot(data=df8,x='Number of Credit Problems',y='Number of Open Accounts')
plt.xlabel('No. of Credit Problems')
plt.ylabel('No. of Open Accounts')
plt.grid()         # Grid Lines for better readability
plt.title('Relationship between Number of Credit Problems and Number of Open Accounts')
plt.show()
No description has been provided for this image
In [124]:
df8[['Number of Credit Problems','Number of Open Accounts']].corr()
Out[124]:
Number of Credit Problems Number of Open Accounts
Number of Credit Problems 1.000000 -0.013995
Number of Open Accounts -0.013995 1.000000

Since the correlation between Number of Credit Problems and Number of Open Accounts is -0.013995, we can say that both of them are not related to each other¶

10) What is the distribution of annual income across different loan statuses?¶

In [127]:
df9 = df.dropna(subset='Annual Income')

# Creating a New DataFrame and removing all the null values
In [128]:
df9['Annual Income'].isna().unique()     

# To check null values
Out[128]:
array([False])
In [129]:
df9['Loan Status'].isna().unique()      

# To check null values
Out[129]:
array([False])

Graphical Representation¶

In [131]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df9,x='Annual Income', y='Loan Status')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Loan Status')
plt.grid(axis='x')    # Grid Lines for better readability
plt.title('Distribution of Annual Income & Different Loan Status')
plt.show()
No description has been provided for this image
In [132]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df9,x='Annual Income', y='Loan Status')
plt.xlim(-200000,10000000)     # To not include outliers and for better readability
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Loan Status')
plt.grid(axis='x')    # Grid Lines for better readability
plt.title('Distribution of Annual Income & Different Loan Status')
plt.show()
No description has been provided for this image
In [133]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df9,x='Annual Income', y='Loan Status')
plt.xlim(-200000,5000000)     # To not include outliers and for better readability
plt.xlabel('Annual Income (per Million USD)')
plt.ylabel('Loan Status')
plt.grid(axis='x')    # Grid Lines for better readability
plt.title('Distribution of Annual Income & Different Loan Status')
plt.show()
No description has been provided for this image

Numeric Representation¶

In [135]:
df9.groupby('Loan Status')['Annual Income'].describe()
Out[135]:
count mean std min 25% 50% 75% max
Loan Status
Charged Off 17210.0 1.267790e+06 1.506251e+06 76627.0 778743.50 1085812.0 1507578.75 165557393.0
Fully Paid 63636.0 1.408157e+06 9.315770e+05 81092.0 865853.75 1213026.5 1694952.00 36475440.0

11) Is there a correlation between the current loan amount and the number of open accounts?¶

In [137]:
df1['Current Loan Amount'].isna().unique()   

# Using already made up DataFrame and checking null values
Out[137]:
array([False])
In [138]:
df1['Number of Open Accounts'].isna().unique()      

# Using already made up DataFrame and checking null values
Out[138]:
array([False])
In [139]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df1,x='Current Loan Amount', y = 'Number of Open Accounts')
plt.xlabel('Current Loan Amount')
plt.ylabel('No. of Open Accounts')
plt.grid(axis='y')         # Grid Lines for better understanding
plt.title('Correlation Between Current & No. of Open Accounts')
plt.show()
No description has been provided for this image
In [140]:
df1[['Number of Open Accounts','Current Loan Amount']].corr()
Out[140]:
Number of Open Accounts Current Loan Amount
Number of Open Accounts 1.000000 0.203099
Current Loan Amount 0.203099 1.000000

As the Correlation between Number of Open Accounts and Current Loan Amount is 0.203099, the correlation between them is on a lower side¶

12) How does the monthly debt vary with the years of credit history?¶

In [143]:
df10 = df.dropna(subset='Monthly Debt')
In [144]:
df10['Monthly Debt'].isna().unique()

# Using already made up DataFrame and checking null values
Out[144]:
array([False])
In [145]:
df10['Years of Credit History'].isna().unique()     

# Using already made up DataFrame and checking null values
Out[145]:
array([False])
In [146]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df10,x='Monthly Debt', y='Years of Credit History')
plt.grid(axis='x')          # Grid Lines for better readability
plt.xlabel('Monthly Debt')
plt.ylabel('Years of Credit History')
plt.title('Monthly Debt vs Years of Credit')
plt.show()
No description has been provided for this image
In [147]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df10,x='Monthly Debt', y='Years of Credit History')
plt.xlim(-20000,225000)     # To not include outliers
plt.grid(axis='x')         # Grid Lines for better readability
plt.xlabel('Monthly Debt')
plt.ylabel('Years of Credit History')
plt.title('Monthly Debt vs Years of Credit')
plt.show()
No description has been provided for this image
In [148]:
pd.DataFrame(df10[['Monthly Debt','Years of Credit History']].var().round(2))
Out[148]:
0
Monthly Debt 1.482304e+08
Years of Credit History 4.921000e+01
In [149]:
pd.DataFrame(df10[['Monthly Debt','Years of Credit History']].std().round(2))
Out[149]:
0
Monthly Debt 12174.99
Years of Credit History 7.02

13) What is the average annual income for each purpose of loan?¶

Numeric Representation¶

In [152]:
df4
Out[152]:
Annual Income
Purpose
Business Loan 1591962.79
Buy House 1495251.47
Buy a Car 1220352.74
Debt Consolidation 1362463.57
Educational Expenses 1071051.32
Home Improvements 1667437.36
Medical Bills 1409286.17
Other 1348516.49
Take a Trip 1219898.67
major_purchase 1430753.50
moving 1591375.25
other 1286008.49
renewable_energy 1473739.75
small_business 1547456.22
vacation 1155289.53
wedding 1464232.62

Graphical Representation¶

In [154]:
plt.figure(figsize=(10,6))
sns.barplot(data=df4,x='Annual Income',y='Purpose')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Purpose')
plt.grid(axis='x')    # Grid Lines for better readability
plt.title('Purpose of Loans compared to Annual Income')
plt.show()
No description has been provided for this image

14) How does the credit score vary with the number of credit problems?¶

In [156]:
df2['Credit Score'].isna().unique()  

# Using already made up DataFrame and checking null values
Out[156]:
array([False])
In [157]:
df2['Number of Credit Problems'].isna().unique()  

# Using already made up DataFrame and checking null values
Out[157]:
array([False])
In [158]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Credit Score',y='Number of Credit Problems')
plt.xlabel('Credit Score')
plt.ylabel('No. of Credit Problems')
plt.grid()          # Grid Lines for better readability
plt.title('Credit Score vs No. of Credit Problems')
plt.show()
No description has been provided for this image
In [159]:
pd.DataFrame(df2[['Credit Score','Number of Credit Problems']].corr())
Out[159]:
Credit Score Number of Credit Problems
Credit Score 1.00000 -0.04726
Number of Credit Problems -0.04726 1.00000
In [160]:
pd.DataFrame(df2[['Credit Score','Number of Credit Problems']].var().round(2))
Out[160]:
0
Credit Score 800.73
Number of Credit Problems 0.24
In [161]:
pd.DataFrame(df2[['Credit Score','Number of Credit Problems']].std().round(2))
Out[161]:
0
Credit Score 28.30
Number of Credit Problems 0.49

15) Is there a correlation between the number of credit problems and the current credit balance?¶

In [163]:
df6['Current Credit Balance'].isna().unique()    

# Using already made up DataFrame and checking null values
Out[163]:
array([False])
In [164]:
df6['Number of Credit Problems'].isna().unique()    

# Using already made up DataFrame and checking null values
Out[164]:
array([False])
In [165]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df6,x='Number of Credit Problems', y='Current Credit Balance')
plt.xlabel('No. of Credit Balance')
plt.ylabel('Current Credit Balance (per Million USD)')
plt.grid()    # Grid Lines for better readability 
plt.title('Correlation between No. of Credit Problems and Current Credit Balance')
plt.show()
No description has been provided for this image
In [166]:
df6[['Number of Credit Problems','Current Credit Balance']].corr()
Out[166]:
Number of Credit Problems Current Credit Balance
Number of Credit Problems 1.000000 -0.163361
Current Credit Balance -0.163361 1.000000

As the Correlation between Number of Credit Problems and Current Credit Balance is -0.163361, the correlation between them is on a weaker side¶

16) What is the distribution of current loan amounts across different home ownership types?¶

In [169]:
df1['Current Loan Amount'].isna().unique()     

# Using already made up DataFrame and checking null values
Out[169]:
array([False])
In [170]:
df1['Home Ownership'].isna().unique()     

# Using already made up DataFrame and checking null values
Out[170]:
array([False])

Graphical Representation¶

In [172]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df1,x='Current Loan Amount',y='Home Ownership')
plt.xlabel('Current Loan Amount')
plt.ylabel('Home Ownership')
plt.grid(axis='x')      # Gridlines for better readability
plt.title('Distribution of Current Loan Amount across type of Home Ownership')
plt.show()
No description has been provided for this image

Numeric Representation¶

In [174]:
df1.groupby('Home Ownership')['Current Loan Amount'].describe()
Out[174]:
count mean std min 25% 50% 75% max
Home Ownership
HaveMortgage 185.0 127574.789189 106018.905122 10802.0 55858.0 108350.0 155826.0 733106.0
Home Mortgage 42690.0 350595.067323 193998.628060 21472.0 202268.0 324434.0 469249.0 789250.0
Own Home 8148.0 298699.116593 184059.002124 21472.0 155683.0 259336.0 412764.0 789030.0
Rent 37493.0 272596.091750 167881.641125 11242.0 145178.0 226952.0 357742.0 789250.0

17) How does the annual income vary with the years in the current job?¶

In [176]:
df11 = df.dropna(subset='Annual Income')

# Creating a New DataFrame and removing all the null values
In [177]:
df11['Annual Income'].isna().unique()

# To check null values
Out[177]:
array([False])
In [178]:
df11['Years in current job'].isna().unique()

# To check null values
Out[178]:
array([False,  True])
In [179]:
df11 = df11.dropna(subset='Years in current job')   

# Removing null Values
In [180]:
df11['Years in current job'].unique()       

# To recheck null Values
Out[180]:
array(['8 years', '3 years', '10+ years', '< 1 year', '2 years',
       '4 years', '1 year', '6 years', '5 years', '7 years', '9 years'],
      dtype=object)
In [181]:
# Apply the conversion function to the "Years in current job" column

df11['Years in current job'] = df11['Years in current job'].apply(convert_to_integers)
In [182]:
df11['Years in current job'].unique()  

# To recheck null Values
Out[182]:
array([ 8,  3, 10,  0,  2,  4,  1,  6,  5,  7,  9], dtype=int64)
In [183]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df11,x='Annual Income', y='Years in current job')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Years in Current Job')
plt.grid()       # Grid Lines for better readability   
plt.title('Annual Income vs Years in Current Job')
plt.show()
No description has been provided for this image
In [184]:
df11[['Years in current job','Annual Income']].corr()
Out[184]:
Years in current job Annual Income
Years in current job 1.000000 0.078364
Annual Income 0.078364 1.000000
In [185]:
df11 = df11[df11['Annual Income'] < 50000000]    

# To remove Outliers
In [186]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df11,x='Annual Income', y='Years in current job')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Years in Current Job')
plt.grid()          # Grid Lines for better readability
plt.title('Annual Income vs Years in Current Job')
plt.show()
No description has been provided for this image
In [187]:
df11[['Years in current job','Annual Income']].corr()
Out[187]:
Years in current job Annual Income
Years in current job 1.000000 0.090432
Annual Income 0.090432 1.000000
In [188]:
df11 = df11[df11['Annual Income'] < 20000000]    

# To remove Outliers
In [189]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df11,x='Annual Income', y='Years in current job')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Years in Current Job')
plt.grid()       # Grid Lines for better readability
plt.title('Annual Income vs Years in Current Job')
plt.show()
No description has been provided for this image
In [190]:
df11[['Years in current job','Annual Income']].corr()
Out[190]:
Years in current job Annual Income
Years in current job 1.000000 0.093857
Annual Income 0.093857 1.000000

As the Correlation between Number of Credit Problems and Current Credit Balance is 0.093857, the correlation between them is negligible¶

18) Is there a correlation between the current loan amount and the monthly debt?¶

In [193]:
df1['Current Loan Amount'].isna().unique()

# Using already made up DataFrame and checking null values
Out[193]:
array([False])
In [194]:
df1['Monthly Debt'].isna().unique()

# Using already made up DataFrame and checking null values
Out[194]:
array([False])
In [195]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df1,x='Monthly Debt',y='Current Loan Amount')
plt.xlabel('Monthly Debt')
plt.ylabel('Current Loan Amount')
plt.grid()    # Grid Lines for better readability
plt.title('Relationship between Current Loan Amount and Monthly Debt')
plt.show()
No description has been provided for this image
In [196]:
df1[['Current Loan Amount','Monthly Debt']].corr()
Out[196]:
Current Loan Amount Monthly Debt
Current Loan Amount 1.000000 0.438728
Monthly Debt 0.438728 1.000000
In [197]:
df12 = df1[df1['Monthly Debt'] < 200000]   

# To remove Outliers
In [198]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df12,x='Monthly Debt',y='Current Loan Amount')
plt.xlabel('Monthly Debt')
plt.ylabel('Current Loan Amount')
plt.grid()   # Grid Lines for better readability
plt.title('Relationship between Current Loan Amount and Monthly Debt')
plt.show()
No description has been provided for this image
In [199]:
df12[['Current Loan Amount','Monthly Debt']].corr()
Out[199]:
Current Loan Amount Monthly Debt
Current Loan Amount 1.000000 0.442072
Monthly Debt 0.442072 1.000000
In [200]:
df12 = df12[df12['Monthly Debt'] < 90000]

# To remove Outliers again
In [201]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df12,x='Monthly Debt',y='Current Loan Amount')
plt.xlabel('Monthly Debt')
plt.ylabel('Current Loan Amount')
plt.grid()   # Grid Lines for better readability
plt.title('Relationship between Current Loan Amount and Monthly Debt')
plt.show()
No description has been provided for this image
In [202]:
df12[['Current Loan Amount','Monthly Debt']].corr()
Out[202]:
Current Loan Amount Monthly Debt
Current Loan Amount 1.000000 0.446079
Monthly Debt 0.446079 1.000000

As the Correlation between Monthly Debt and Current Loan Amount is 0.446079, both of them are moderately correlated.¶

19) What is the average monthly debt for each home ownership type?¶

In [205]:
df10['Monthly Debt'].isna().unique()    

# Using already made up DataFrame and checking null values
Out[205]:
array([False])
In [206]:
df10['Home Ownership'].isna().unique()

# Using already made up DataFrame and checking null values
Out[206]:
array([False])
In [207]:
df13 = pd.DataFrame(df10.groupby(by='Home Ownership')['Monthly Debt'].mean().round(2))

Numeric Representation¶

In [209]:
df13
Out[209]:
Monthly Debt
Home Ownership
HaveMortgage 18541.78
Home Mortgage 21072.06
Own Home 17060.08
Rent 15796.78

Graphical Representation¶

In [211]:
plt.figure(figsize=(10,6))
sns.barplot(data=df13,x='Home Ownership',y='Monthly Debt')
plt.xlabel('Home Ownership')
plt.ylabel('Avg Monthly Debt')
plt.grid(axis='y')   # Grid Lines for better readability
plt.title('Avg Monthly Debt of Each Ownership Type')
plt.show()
No description has been provided for this image

20) How does the credit score vary with the number of open accounts?¶

In [213]:
df2['Credit Score'].isna().unique()

# Using already made up DataFrame and checking null values
Out[213]:
array([False])
In [214]:
df2['Number of Credit Problems'].isna().unique()

# Using already made up DataFrame and checking null values
Out[214]:
array([False])
In [215]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Credit Score',y='Number of Open Accounts')
plt.xlabel('Credit Score')
plt.ylabel('No. of Open Accounts')
plt.grid()   # Grid Lines for better readability
plt.title('Credit Score vs No. of Open Accounts')
plt.show()
No description has been provided for this image
In [216]:
df2[['Credit Score','Number of Open Accounts']].corr()
Out[216]:
Credit Score Number of Open Accounts
Credit Score 1.000000 -0.030063
Number of Open Accounts -0.030063 1.000000
In [217]:
pd.DataFrame(df2[['Credit Score','Number of Open Accounts']].var().round(2))
Out[217]:
0
Credit Score 800.73
Number of Open Accounts 25.29
In [218]:
pd.DataFrame(df2[['Credit Score','Number of Open Accounts']].std().round(2))
Out[218]:
0
Credit Score 28.30
Number of Open Accounts 5.03

21) What is the distribution of credit scores across different loan statuses?¶

In [220]:
df2['Credit Score'].isna().unique()

# Using already made up DataFrame and checking null values
Out[220]:
array([False])
In [221]:
df2['Loan Status'].isna().unique()

# Using already made up DataFrame and checking null values
Out[221]:
array([False])

Graphical Representation¶

In [223]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df2,x='Credit Score',y='Loan Status')
plt.xlabel('Credit Score')
plt.ylabel('Loan Status')
plt.grid(axis='x')     # Grid Lines for better readability
plt.title('Distribution of Credit Score across different types Loan Status')
plt.show()
No description has been provided for this image

Numeric Representation¶

In [225]:
df2.groupby('Loan Status')['Credit Score'].describe()
Out[225]:
count mean std min 25% 50% 75% max
Loan Status
Charged Off 17210.0 710.390471 31.26153 585.0 695.0 719.0 734.0 751.0
Fully Paid 63636.0 717.889874 27.22175 585.0 705.0 723.0 739.0 751.0

22) Is there a correlation between the current loan amount and the years of credit history?¶

In [227]:
df1['Current Loan Amount'].isna().unique()
Out[227]:
array([False])
In [228]:
df1['Years of Credit History'].isna().unique()
Out[228]:
array([False])
In [229]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df1,x='Current Loan Amount',y='Years of Credit History')
plt.xlabel('Current Loan Amount')
plt.ylabel('Years of Credit History')
plt.grid()  # Grid Lines for better readability
plt.title('Relationship between Current Loan Amount & Years of Credit History')
plt.show()
No description has been provided for this image
In [230]:
df1[['Current Loan Amount','Years of Credit History']].corr()
Out[230]:
Current Loan Amount Years of Credit History
Current Loan Amount 1.000000 0.157962
Years of Credit History 0.157962 1.000000

As the Correlation between Years of Credit History and Current Loan Amount is 0.157962, the correlation between them is very low¶

23) How does the monthly debt vary with the number of credit problems?¶

In [233]:
df10['Monthly Debt'].isna().unique()

# Using already made up DataFrame and checking null values
Out[233]:
array([False])
In [234]:
df10['Number of Credit Problems'].isna().unique()

# Using already made up DataFrame and checking null values
Out[234]:
array([False])
In [235]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df10,x='Monthly Debt',y='Number of Credit Problems')
plt.xlabel('Monthly Debt')
plt.ylabel('No. of Credit Problems')
plt.grid()   # Grid Lines for better readability
plt.title('Monthly Debt vs No. of Credit Problems')
plt.show()
No description has been provided for this image
In [236]:
df10[['Monthly Debt','Number of Credit Problems']].corr()
Out[236]:
Monthly Debt Number of Credit Problems
Monthly Debt 1.000000 -0.055383
Number of Credit Problems -0.055383 1.000000
In [237]:
pd.DataFrame(df10[['Monthly Debt','Number of Credit Problems']].var().round(2))
Out[237]:
0
Monthly Debt 1.482304e+08
Number of Credit Problems 2.300000e-01
In [238]:
pd.DataFrame(df10[['Monthly Debt','Number of Credit Problems']].std().round(2))
Out[238]:
0
Monthly Debt 12174.99
Number of Credit Problems 0.48

24) What is the average current loan amount for each purpose of loan?¶

In [240]:
df['Purpose'].isna().unique()

# To check null values
Out[240]:
array([False,  True])
In [241]:
df14 = df.dropna(subset='Purpose')

# Creating a New DataFrame and removing all the null values
In [242]:
df14['Purpose'].isna().unique()

# To check null values
Out[242]:
array([False])
In [243]:
df14['Current Loan Amount'].isna().unique()

# To check null values
Out[243]:
array([False])
In [244]:
df14 = df14.groupby(by='Purpose')['Current Loan Amount'].mean().round(2).reset_index()

Numeric Representation¶

In [246]:
df14
Out[246]:
Purpose Current Loan Amount
0 Business Loan 317074.92
1 Buy House 293399.63
2 Buy a Car 154720.50
3 Debt Consolidation 290887.46
4 Educational Expenses 143941.33
5 Home Improvements 275927.38
6 Medical Bills 175237.69
7 Other 197810.18
8 Take a Trip 121905.53
9 major_purchase 205468.31
10 moving 167388.76
11 other 189910.00
12 renewable_energy 197553.40
13 small_business 296953.20
14 vacation 127987.07
15 wedding 211687.25

Graphical Representation¶

In [248]:
plt.figure(figsize=(10,6))
sns.barplot(data=df14,x='Current Loan Amount',y='Purpose')
plt.xlabel('Avg Current Loan Amount')
plt.ylabel('Purpose')
plt.grid(axis='x')  # Grid lines for better readability
plt.title('Average Current Loan Amount for each Loan Status')
plt.show()
No description has been provided for this image

25) How does the credit score vary with the current credit balance?¶

In [250]:
df2['Credit Score'].isna().unique()

# Using already made up DataFrame and checking null values
Out[250]:
array([False])
In [251]:
df2['Current Credit Balance'].isna().unique()

# Using already made up DataFrame and checking null values
Out[251]:
array([False])
In [252]:
df2['Current Credit Balance'].max()

# To check the range
Out[252]:
32878968.0
In [253]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Credit Score',y='Current Credit Balance')
plt.xlabel('Credit Score')
plt.ylabel('Current Credit Balance (per 10 Million USD)')
plt.grid()   # Grid Lines for better readability
plt.title('Credit Score vs Current Credit Balance')
plt.show()
No description has been provided for this image
In [254]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Credit Score',y='Current Credit Balance')
plt.xlabel('Credit Score')
plt.ylabel('Current Credit Balance (per Million USD)')
plt.ylim(-100000,7500000)  # To not include outliers
plt.grid()    # Grid Lines for better readability
plt.title('Credit Score vs Current Credit Balance')
plt.show()
No description has been provided for this image
In [255]:
df2[['Credit Score','Current Credit Balance']].corr()
Out[255]:
Credit Score Current Credit Balance
Credit Score 1.000000 -0.015964
Current Credit Balance -0.015964 1.000000
In [256]:
pd.DataFrame(df2[['Credit Score','Current Credit Balance']].var().round(2))
Out[256]:
0
Credit Score 8.007300e+02
Current Credit Balance 1.456756e+11
In [257]:
pd.DataFrame(df2[['Credit Score','Current Credit Balance']].std().round(2))
Out[257]:
0
Credit Score 28.30
Current Credit Balance 381674.71

26) Is there a correlation between the annual income and the current credit balance?¶

In [259]:
df6['Current Credit Balance'].isna().unique()

# Using already made up DataFrame and checking null values
Out[259]:
array([False])
In [260]:
df6['Annual Income'].isna().unique()

# Using already made up DataFrame and checking null values
Out[260]:
array([False,  True])
In [261]:
df6= df6.dropna(subset='Annual Income')

# To remove outliers
In [262]:
df6['Annual Income'].isna().unique()

# To recheck outliers
Out[262]:
array([False])
In [263]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df6,x='Annual Income',y='Current Credit Balance')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Current Credit Balance (per 10 Million USD)')
plt.grid()   # Grid Lines for better readability
plt.title('Relationship between Annual Income & Current Credit Balance')
plt.show()
No description has been provided for this image
In [264]:
df6[['Annual Income','Current Credit Balance']].corr()
Out[264]:
Annual Income Current Credit Balance
Annual Income 1.000000 0.256798
Current Credit Balance 0.256798 1.000000
In [265]:
df6['Current Credit Balance'].max()    

# To check range
Out[265]:
999400.0
In [266]:
df15 = df6[df6['Annual Income'] < 20000000]   

# To remove Outliers
In [267]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df15,x='Annual Income',y='Current Credit Balance')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Current Credit Balance (per 10 Million USD)')
plt.grid()    # Grid Lines for better readability
plt.title('Relationship between Annual Income & Current Credit Balance')
plt.show()
No description has been provided for this image
In [268]:
df15[['Annual Income','Current Credit Balance']].corr()
Out[268]:
Annual Income Current Credit Balance
Annual Income 1.000000 0.324571
Current Credit Balance 0.324571 1.000000

As the Correlation between Annual Income and Current Credit Balance is 0.324571, the correlation between them is on a weaker side.¶

27) What is the distribution of annual income across different terms (short-term vs. long-term)?¶

In [271]:
df9['Annual Income'].isna().unique()

# Using already made up DataFrame and checking null values
Out[271]:
array([False])
In [272]:
df9['Term'].isna().unique()

# Using already made up DataFrame and checking null values
Out[272]:
array([False])

Graphical Representation¶

In [274]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df9,x='Annual Income',y='Term')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Loan Term')
plt.grid(axis='x')    # Grid Lines for better readability
plt.title('Distribution of Annual Income Across Different Terms')
plt.show()
No description has been provided for this image
In [275]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df9,x='Annual Income',y='Term')
plt.xlabel('Annual Income (per 10 Million USD)')
plt.ylabel('Loan Term')
plt.xlim(-20000,10000000)    # To not include outliers     
plt.grid(axis='x')        # Grid Lines for better readability
plt.title('Distribution of Annual Income Across Different Terms')
plt.show()
No description has been provided for this image
In [276]:
plt.figure(figsize=(10,6))
sns.boxplot(data=df9,x='Annual Income',y='Term')
plt.xlabel('Annual Income (per Million USD)')
plt.ylabel('Loan Term')
plt.xlim(-20000,5000000)     # To not include outliers
plt.grid(axis='x')       # Grid Lines for better readability
plt.title('Distribution of Annual Income Across Different Terms')
plt.show()
No description has been provided for this image

Numeric Representation¶

In [278]:
df9.groupby('Home Ownership')['Current Loan Amount'].describe()
Out[278]:
count mean std min 25% 50% 75% max
Home Ownership
HaveMortgage 185.0 104680.400000 98913.199921 0.0 33770.0 89892.0 133848.0 733106.0
Home Mortgage 39089.0 301853.723196 219598.867028 0.0 131428.0 268972.0 445236.0 789250.0
Own Home 7480.0 257700.758824 200350.536967 0.0 109378.5 221232.0 375694.0 789030.0
Rent 34092.0 236239.467910 182991.019146 0.0 105374.5 215754.0 335852.0 789250.0

28) How does the credit score vary with the number of credit problems?¶

In [280]:
df2['Credit Score'].isna().unique()

# Using already made up DataFrame and checking null values
Out[280]:
array([False])
In [281]:
df2['Number of Credit Problems'].isna().unique()

# Using already made up DataFrame and checking null values
Out[281]:
array([False])
In [282]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df2,x='Credit Score',y='Number of Credit Problems')
plt.xlabel('Credit Score')
plt.ylabel('No. of Credit Problems')
plt.grid()  # Grid Lines for better readability
plt.title('Credit Score vs No. of Credit Problems')
plt.show()
No description has been provided for this image
In [283]:
df2[['Credit Score','Number of Credit Problems']].corr()
Out[283]:
Credit Score Number of Credit Problems
Credit Score 1.00000 -0.04726
Number of Credit Problems -0.04726 1.00000
In [284]:
pd.DataFrame(df2[['Credit Score','Number of Credit Problems']].var().round(2))
Out[284]:
0
Credit Score 800.73
Number of Credit Problems 0.24
In [285]:
pd.DataFrame(df2[['Credit Score','Number of Credit Problems']].std().round(2))
Out[285]:
0
Credit Score 28.30
Number of Credit Problems 0.49

29) Is there a correlation between the current loan amount and the number of credit problems?¶

In [287]:
df1['Current Credit Balance'].isna().unique()

# Using already made up DataFrame and checking null values
Out[287]:
array([False])
In [288]:
df1['Number of Credit Problems'].isna().unique()

# Using already made up DataFrame and checking null values
Out[288]:
array([False])
In [289]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df1,x='Current Loan Amount',y='Number of Credit Problems')
plt.xlabel('Current Loan Amount')
plt.ylabel('No. of Credit Problems')
plt.grid()      # Grid Lines for better readability
plt.title('Current Loan Amount vs No. of Credit Problems')
plt.show()
No description has been provided for this image
In [290]:
df1[['Current Loan Amount','Number of Credit Problems']].corr()
Out[290]:
Current Loan Amount Number of Credit Problems
Current Loan Amount 1.000000 -0.076698
Number of Credit Problems -0.076698 1.000000

As the Correlation between Number of Credit Problems and Current Loan Amount is -0.076698, the correlation between them is negligible¶

30) What is the relationship between the number of open accounts and the years of credit history?¶

In [293]:
df3['Number of Open Accounts'].isna().unique()

# Using already made up DataFrame and checking null values
Out[293]:
array([False])
In [294]:
df3['Years of Credit History'].isna().unique()

# Using already made up DataFrame and checking null values
Out[294]:
array([False])
In [295]:
plt.figure(figsize=(10,6))
sns.scatterplot(data=df3,x='Number of Open Accounts',y='Years of Credit History')
plt.xlabel('No. of Open Accounts')
plt.ylabel('Years of Credit History')
plt.grid()   # Grid Lines for better readability
plt.title('Relationship between No. of Open Accounts vs Years of Credit History')
plt.show()
No description has been provided for this image
In [296]:
df3[['Number of Open Accounts','Years of Credit History']].corr()
Out[296]:
Number of Open Accounts Years of Credit History
Number of Open Accounts 1.000000 0.132147
Years of Credit History 0.132147 1.000000

As the Correlation between No. of Open Accounts and Years of Credit History is 0.132349, the correlation between them is negligible¶

In [ ]: